VCC: Efficient Voxel-Based Collision Checking Framework for Real-Time Robotic Motion Planning
Ching Chen, TSUNG TAI YEH
AI summary
Problem
The pre-planning stage, which involves point cloud filtering and environment data structure construction, consumes excessive computational time, creating a bottleneck that limits a robot's real-time responsiveness to dynamic obstacles.
Approach
VCC optimizes the planning pipeline by filtering redundant point cloud data, partitioning the workspace into an adaptive voxel grid, and organizing the data into a sparse, SIMD-aligned Multilevel Voxel Table for fast collision queries.
Key results
- 3.63× speedup in point cloud filtering
- 220.48× faster data structure construction with 97.73% memory reduction
- 1.94× acceleration of sampling-based motion planning
- 7.71× end-to-end speedup on desktop CPUs and 4.23× on embedded hardware
Why it matters
It makes real-time, collision-free navigation practical for edge-deployed robots by eliminating the computational bottlenecks that previously hindered dynamic obstacle avoidance.
Abstract
To navigate environments with dynamic obstacles, a robot must continuously scan for them and find collision- free paths to reach a goal position. This process starts with receiving obstacle information in the form of a point cloud, followed by a pre-planning stage that involves preprocessing to remove unnecessary points and constructing an environment data structure. However, the pre-planning stage can consume more than 16× the runtime of the planning stage, slowing the robot’s reaction speed. Thus, in this work, we propose VCC, an efficient collision checking framework that primarily targets the pre-planning bottleneck. VCC first cleans the point cloud using Center-selective Voxel Filtering. It then divides the environment into voxels using Adaptive Workspace Voxelization and orga- nizes them in a Multilevel Voxel Table (MVT). In addition, VCC manages the MVT in two memory pools to ensure high data locality and SIMD-aligned data layout. During motion planning, the planner can perform low-latency SIMD-accelerated collision checking using the MVT. Compared with the state-of-the-art method, the experimental results show a 3.63× speedup in filtering. In terms of environment data structure, MVT achieves a 220.48× speedup during construction and reduces memory usage by 97.73%. Additionally, VCC accelerates sampling- based planning by 1.94×. Altogether, VCC achieves an end-to- end speedup of 7.71× on the desktop CPU platform and 4.23× on the embedded computer platform, making real-time motion planning practical for resource-constrained edge devices.